Age of Information Minimization in Hierarchical Aerial Networks via MADDPG Approach
摘要
This work introduces a hierarchical aerial computing system designed to minimize the Age of Information (AoI) for ground devices. The system utilizes a high-altitude platform in conjunction with multiple unmanned aerial vehicles to provide edge computing capabilities. Non-orthogonal multiple access is employed for uplink communication to improve efficiency. To achieve minimal average AoI, a multi-agent deep deterministic policy gradient (MADDPG)-based approach is proposed. This approach optimizes the trajectories of the aerial vehicles, their transmit power, and the allocation of computing resources. Simulation results validate the feasibility of the proposed system.